Quantitative Methods for Strategic and Investment Planning in the Oil-Refining Industry

Transcription

1 Quantitative Methods for Strategic and Investment Planning in the Oil-Refining Industry Brenno C. Menezes, Ignacio E. Grossmann, Lincoln F. Moro and Jeffrey D. Kelly PETROBRAS Current Tool for Strategic Planning (PLANINV) LP No Process Design Synthesis Quantitative Methods Only optimize streams transfers (oil and fuels import/export, market supply) PLANINV Process Design Opt. (MILP) + NLP Processing Blending Goal: develop high-level decision-making optimization to predict structural modifications in refining and logistics assets using more rigorous formulations What, Where, When to Invest? Atmospheric Distillation Delayed Coker 1 Terminal/Pipeline

2 Capital Investment Planning Formulation Maximize: NPV = DemandSales - SupplyCosts - OperatingCosts - InvestmentCosts Subject to: QQQQ rr,uu,nn,tt = EEEEEEEEEE rr,uu,nn + QQQQ rr,uu,nn,tt 11 + QQQQ rr,uu,nn,tt 11 eeeeeeeeeeeeeeeeee: uu, nn eeeeee Where: QQQQ rr,uu,nn,tt = QQQQ rr,uu,nn,tt 11 + QQQQ rr,uu,nn,tt 11 iiiiiiiiiiiiiiiiiiiiiiii: uu, nn iiiiii yyyy rr,uu,nn,tt QQQQ uu LL QQQQ rr,uu,nn,tt yyyy rr,uu,nn,tt QQQQ uu UU yyyy rr,uu,nn,tt QQQQ uu LL QQQQ rr,uu,nn,tt yyyy rr,uu,nn,tt QQQQ uu UU QQQQ rr,uu,nn,tt QQQQ rr,uu,nn,tt QF= operational flow QC= total capacity yyyy=expansion of an existent unit yyi= installation of a new unit uu, nn eeeeee uu, nn iiiiii QEE= expanded capacity QII= installed capacity Sahinidis et. al., CACE, 13, (1989) and Sahinidis & Grossmann, CACE, 15, (1991). (R,U,N,T) R=Refinery U=Unit type N=Number of an unit type T=Time INVREF QC=QC t-1 +Q NEW MILP OPREF QF QC NLP Crude dieting Improved Formulation: - Installations; - Project Execution; - NLP Operational Layer. Project execution T1 Take an investment decision (binary) T2 year Count on the additional production Processing Blending ISW ON

3 Options to Formulate the Problem Full Space Problem Investment MINLP MILP QF u,t QC u,t link constraint Operational NLP Crude Diet Processing Blending - Crude - Cuts/Final Cuts - Final Products 1 st - NLP Operational Problem Z=profit (m 3 /d) and QF u =unit throughputs to control capacity expansion 2 nd - MINLP Investments Problem (NLP Operational Problem Embedded) Z=NPV ($) and QE u,t and QC u,t to control capacity expansion Aggregated Multi-Site Approach 3 rd - MILP Investments Problem + NLP Operational Problem (Phenomenological Decomposition Heuristics) Z=NPV ($) and QE u,t, QI u,t and QC u,t to control capacity expansion and installation Multi-Site Approach (1) Menezes, et. al. (2014). Nonlinear Production Planning of Oil-Refinery Units for the Future Fuel Market in Brazil: Process Design Scenario- Based Model. Ind. Eng. Chem. Res. DOI: /ie Published online: Feb, 17,

4 Aggregated Multi-Site Approach Conceptual Projects under reevaluation Refining Process Capacities (k m 3 /d) Crude Distillation Unit CDU Vacuum Distillation Unit VDU Residue Fluid Catalytic Cracking RFCC Fluid Catalytic Cracking FCC Hydrocracking HCC Propane Deasphalting PDA Delayed Coker DC Light Cracked Naphtha Hydrotreater LCNHT Coker Light Naphtha Hydrotreater CLNHT Stabilizer ST Kerosene Hydrotreater KHT Diesel Hydrotreater (medium severity) D1HT Diesel Hydrotreater (high severity) D2HT Reformer REF

5 Aggregated Multi-Site Approach Unit (u) α u β u mi US$/1000m 3 mi US$ CDU VDU FCC HCC RFCC DC KHT DHT LCNHT CLNHT ST REF Crude CDU N SW1 K SW2 LD SW3 HD ATR FN FK FLD FHD VDU VR C1C2 C3C4 LVGO HVGO PDA 00 RFCC COKE DAO ASPR DC FCC C1C2 C3C4 CLN CHN CLGO CMGO CHGO HCN LCO DO C1C2 C3C4 LCN LCNHT HTLCN HCC C1C2 C3C4 HCCN HCCK HCCD LNST ST HCCO HNST GOST CLNHT HTCLN C1C2 C3C4 KHT HTK FG REF REFOR LPG PQN LPGimp JET JETimp GLNimp (GLNA) ETH GLN (GLNC) H2 LSDimp LSD MSD HSD Perform linear regression to convert the nonlinear Power Law to Fixed-Charge Relation: Investment Cost = a*capacity+b*setup For RNEST D1HT D2HT 12 discrete variables; 1127 continuous variables 1019 equations 4463 nonzero elements; 2552 nonlinear elements max NPV = 1 tr F t0 1 (1 + ir) t (pr p,t 0 0 Refc t0 )Dem p,t p t t 0 pr cr,t0 υ cr,t QF CDU,t pr imp,t0 QF imp,t cr HTD imp USD COKE FO Υ HT,t QF HT,t HT 1 (1 + ir) t it (α u,t it QE u,t + β u,tit y u,t ) u t<t end 5

6 Aggregated Multi-Site Approach 2020 Expansions, QE (k m 3 /d) trends 4.2% p.a. QF u,t M (=1000) CONOPT DICOPT (NLP: CONOPT) 2016 Unit GLNC GLNC ETH GLNC GLNC ETH 372 CDU VDU RFCC FCC HCC DC KHT D2HT LCNHT CLNHT ST REF Profit (mi US$/d) CPU (s) Expansions, QE (k m 3 /d) trends 4.2% p.a Unit GLNC GLNC ETH GLNC GLNC ETH 372 CDU VDU RFCC FCC HCC DC KHT D2HT LCNHT CLNHT ST REF NPV (bi US$) Investment (bi US$) Profit (mi US$/d) CPU (s) Operational Crude Diet Processing Blending - Crude - Cuts/Final Cuts - Final Products Full Space Problem Investment Operational Crude Diet Processing NLP Blending - Crude - Cuts/Final Cuts - Final Products MINLP MILP QF u,t QC u,t link constraint NLP 6

7 Multi-Site Approach PDH (MILP+NLP) PDH = Phenomenological Decomposition Heuristics = Quantity + Quality problems Decomposed Quantity Problem: Logic + Quantity variables Quality Problem: Quantity + Quality Variables MILP NLP 7

8 REPLAN CDU.(1,2) VDU.(1,2) FCC.(1,2) DC.(1,2) LCNHT.(1,2) CLNHT.(1,2) DHT.(1,2) REF.1 15 units U.N.T U.N.T U.N.T (20 exp / 3 inst) CDU.2.2 VDU.1.2 FCC.1.1 DHT.2.1 LCNHT.1.1 CLNHT.1.1 CDU.3.1 VDU expans 2 install REVAP CDU.1 VDU.1 FCC.1 PDA.1 DC.1 LCNHT.1 CLNHT.1 KHT.(1,2) DHT.(1,2) REF.1 12 units CDU.2.1 VDU.1.1 FCC.1.1 LCNHT.1.1 CLNHT.1.1 KHT.1.1 KHT.2.2 DHT.2.1 CDU expans 1 install MILP/NLP (T=3)* step 1 2 CPU (s) MILP (CPLEX) NPV (bi U$) NLP (CONOPT) infeas NLP (SNOPT) NLP (IPOPT) MILP (CPLEX) NLP (CONOPT) infeas NLP (SNOPT) NLP (IPOPT) RPBC CDU.(1,2,3) VDU.(1,2) FCC.1 DC.(1,2) LCNHT.(1,2) CLNHT.1 DHT.(1,2) REF.1 ALK.1 15 units NLP (T=1) MILP (T=3) NLP (T=3) CDU.1.1 VDU.1.1 FCC.1.1 LCNHT.1.1 CLNHT.1.1 DHT expans Discrete Var. Intel i7-3820qm 2.7GHz 16GB (R,U,N,T) Refinery Unit Type Number of the Unit type Time Eq. Var. Nonzero Non- Linear --- 1,538 1,986 14,888 9, ,960 7,620 28, ,223 10,751 70,376 45,446 *MINLP (DICOPT) problem does not converge due to NLP solver infeasibilities 8

9 Conclusions Novelty: Aggregated multi-site approach for capacity expansion of a country/company Nonlinearities from processing and blending to evaluate the capability Includes project execution time (excluding the production from expanded units during this period) Expansion and Installation to control the capacity increment of units Phenomenological decomposition (quantity + quality problems segregated) More realistic approach (in a quantitative manner) for strategic and investment planning in the oil-refining industry 9

10 Conclusions Impact for industrial applications: Aggregated model (NLP and MINLP cases) used to define the overall capacity increment per type of oil-refinery unit demanded in the conceptual projects Realistic formulation to predict investments in oil-refinery units Avoids overestimating/underestimating capacity expansion/installation Evaluates the capability (not only the capacity) by including nonlinearities 10